Rembrand CEO Omar Tawakol - why product placement works better than traditional ads

Omar Tawakol is the CEO and founder of Rembrand, an AI-powered virtual product placement company. He founded BlueKai, a data management platform acquired by Oracle for $400M+ in 2014, and Voicera, an AI meeting assistant acquired by Cisco in 2019. He currently serves on the board of The Trade Desk and as a General Partner at super{set}. In this episode of World of DaaS, Omar and Auren discuss: Why virtual billboards work better than 3D product placementHow AI eliminates two-year product place...

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:02Auren Hoffman?

  Hello fellow data nerds. Welcome to World of DAS. I'm your host, Orin Hoffman, CEO of Incubate and GP of Flex Capital. Discover more episodes, get weekly data as a service news, original content, articles on data and more at worldofdas.com. That's worldofdas, D-A-A-S, dot com. Hello fellow data nerds. My guest today is Omar Taral. Omar is the CEO and founder of Rembrandt, an AI virtual product placement company. He also recently joined Superset as a general partner. He founded BlueKai, which was a data management platform acquired by Oracle for $400 million-ish in 2014, and Voicera, an AI meeting assistant acquired by Cisco

• 0:47Auren Hoffman?

  in 2019. And he currently serves on the board of the Trade Desk. He's also an ad tech OG. So Omar, welcome to World of DAS.

• 0:56Speaker 01

  It's so exciting to be here. Good to see you, Orin.

• 0:58Auren Hoffman?

  We've known each other for a long time. We've spent a lot of time working together back when you were at BlueKai, when I was at LiveRamp. Now you're involved in like the product placement industry. And back in the day, you know, you would like pay to get like, I don't know, Reese's Pieces for ET or something like that. Now I assume you can like AIify it and make it so that every person watching something can see something different. How is that evolving?

• 1:24Speaker 01

  No, it's a fascinating category. And I want to step back for a minute and look at something that Terry Quaj just said recently at this Ad CP event. He said that if you look at the orders of magnitude changes in advertising, they all came from native. So if you look at like search, before we came up with sponsored links, people were trying banner ads on the top of, you know, search pages. It didn't work. And then once they came up with the idea of an ad that was integrated into the search or the sponsored link, that industry took off. And then same thing with like Facebook, early days of Facebook, people were like, yeah, this inventory is low value. Who's going to find value in it? How are they going to make it work? And then they came up with a native format and sponsored posts.

• 2:09Speaker 01

  So if you look at video, which is the majority of people's time spent in consumption, there is a native format for it. It's called product placement. It's existed for decades, but it is so damn manual. It's like two years before a script comes out, you negotiate to get into it.

• 2:24Auren Hoffman?

  To me, like what I like about the product placement as opposed to the interstitial like is that those interstitial ads are really annoying. They take you away from the content. You hate it. A lot of people even pay like me paid to not see them. That's how annoying they are. So we paid to not see those ads. Whereas like if you have a product placement, it could actually be part of the content. In fact, that could even be like real additive to the content.

• 2:49Speaker 01

  Actually, you're hitting the exact core issue, which is people are paying to either avoid ads or get low ad load environments. And this shift is so massive that 10 years ago, people spent 80% of their time in basically ad loads of 17 minutes per hour, which to me feels like torture. It's insane. And people like me and you, actually, if you look at ad execs, almost every ad exec I know literally pays for zero ads. So which is really kind of funny.

• 3:17Auren Hoffman?

  I like ads, like I love the ads on Instagram or something because they're actually useful and they drive behavior in a good way. But like bad ads, I don't want to see. And then of course, ads that are like, make it so it's hard to see the content also is bad too.

• 3:34Speaker 01

  Great. So then you look at like virtual product placement. It gives you this opportunity for you to put something that's non-interruptive, that sits in the environment, that fits into it. And so it's very powerful. And by the way, you could also use it in conjunction with advertising. So for example, you're in a show, you see a Pepsi sitting on the table. And then in the ad, they can tell you why Pepsi. There's this ability to also combine them that makes it like super powerful. You also noted something interesting, which is with the advent of programmatic. We've known how to do it, but just now starting to be capable of doing it where you would see Perrier

• 4:15Speaker 01

  and I would see Pepsi, depending upon who they're targeting, where that's really where you start to have this massive unlock. And we figured that out and we've integrated it. And I think that's going to be the next wave.

• 4:28Auren Hoffman?

  I would love to just have better video experience, period. So not just for the ads, but where you and I see something slightly different when we're watching a video. What I really want is the next time I watch a Marvel movie, I want to see, you know, when they throw like the meteor and there's like a big crowd in New York City or something, I want to see me and my family like in the crowd being like, whoa, you know, I would a hundred percent pay extra for that. Like, that would be so cool to see that. That seems so easy to do that. So just like getting more customized content would be fun, too.

• 5:03Speaker 01

  Yeah, that is actually a whole other angle you can take it. And I've actually, I won't say which streamer, but I've spoken to a streamer who is playing with simple variants of what we're talking about. It's a fascinating world. Let me tell you what's gone wrong with Virg Product Place. Virg Product Placement has been here for a decade, well before Rembrandt. But the problem is the way people attempted to do it before us, they didn't use AI, they used factories of VFX people. And so the process was too full of friction. You went back and forth between the content owner and the advertiser.

• 5:33Auren Hoffman?

  Oh, so you'd have like the product, then you'd like ship it over to someone, then they would like add it in and then they would ship it back and you like review it?

• 5:41Speaker 01

  Yeah, it had cycles like that that were way too slow. Do you like it like this? Do you like it like that? And the advertiser wants something, the publisher doesn't. So what you need to do is figure out a way to separate supply and demand. So you didn't have this like back and forth. That's number one. You need to make it a software driven, so it's not a team of VFX frame by frame, taking $30,000 in a week. And that's a problem we solved. But the problem that also nobody solved, which wasn't a technical one until we came along, was because with VFX, you can do anything. It's essentially thinking like an artist. And so art doesn't scale that easily. And so because you could do anything, people were too unbounded. And when you're too unbounded and you're thinking,

• 6:25Speaker 01

  you can't standardize and you can't run processes that are really fast. And that's the problem we really wanted to crack. And one of the things that we found that was counterintuitive. Was people were really attracted to finding a special dining table environment and putting a can of something on the table. And what we found out really worked well was something completely different. And that is that you're watching a show and the scene is over and all of a sudden you want to show people that your show is in Chicago. So you pan out to a cityscape. And at that moment, there's no actors or actresses. So you're not constrained by what the director wants to put next to them. And so what you can do is put a massive billboard. And then what do you put on the billboard? You put a message people are going to remember, like Sprite now with Cherry. Whereas if you just put a product on the

• 7:10Speaker 01

  table, people will just see a product. But when you put a message and there's nothing else in the scene.

• 7:15Auren Hoffman?

  And people are already used to the billboards, already used to that ad unit. So that makes a lot of sense.

• 7:20Speaker 01

  It looks like it's there. It looks like it actually is in that cityscape. Most people think it's physically there. And so we found that to have massive message recall. And it also allows us to separate supply and demand because the supplier is like, oh, fine, I'll allow billboard here. There's no conflict with the scene. Whereas if it was in the kitchen scene and a certain actor, they'd be saying, well, does that actor drink? They get all hung up on it. Here you've separated supply and demand and things can move much faster.

• 7:45Auren Hoffman?

  So you have to somehow help them understand, okay, this is what the ad unit looks like. Here's how it works. AI finding a different product is weird because it's like, is that period should be part of the story. They may have had a point in the arc a few times ago where the person says, I hate carbonated water or something. And then like, well, why is the period there?

• 8:07Speaker 01

  Exactly. And we still do that because that's the sexy thing. But what we find when you do the billboard is we can give a very clear job to be done for the advertiser. We basically say, we are focusing on advertiser who have new news. You want to share a functional message, which is, by the way, for CPGs, it's about 40% of their advertising, they're delivering a specific message. So we go after the brand directly who has new news to share. And the message recalls through the roof because you're not skipping the ad. There's nothing else in the scene at that moment. People remember it. So that's kind of the niche that we found. And it's a significant niche. That's what we're trying to scale.

• 8:45Auren Hoffman?

  You do have to then work with the content creators to make sure that they've got some place in the scenes to do this, whether it's, you know, as you said, panning out for Chicago or some other place. Because, you know, not everyone will have that today, but it makes sense to do that.

• 9:01Speaker 01

  That's absolutely right.

• 9:02Auren Hoffman?

  It could be even while the credits are rolling and stuff, or introducing the scene, or, you know, whatever it might be.

• 9:08Speaker 01

  There are types of shows that have transition scenes and establishing shots. We may use AI to enhance those, make them longer, make them better, and do things to help the content owners. I don't want to give away too much, but there's an interesting angle there. And what we've also seen, which is shocking to me, or in this one is really shocking, is virtual product placement has taken off in other countries 10x more than in the US. This is kind of like mobile payments, where mobile payments took off in Africa and Asia before it really became big in the US. We've seen the same thing here, is that in a whole bunch of other countries, we see volumes way higher than in the US. It took me a while to get my mind around.

• 9:51Auren Hoffman?

  Why is that? Like, was it just because they had the lack of different advertising forms? Because, you know, these other countries do have billboards, they've got TV, they've got other places to advertise.

• 10:02Speaker 01

  I think it's because in those other markets, linear was tapping out and CTV hadn't grown as fast. And so if you could provide a better monetization without adding ad loan to linear, that's where it really took off. Where we're starting to see kernels of this in the US is when we give the tools to publishers for them to do their own virtual product placement as opposed to sell it for them. Because if you sell it for them, they're not going to give you your best shows, and they're not going to want you to disclose to the advertiser all the details of what shows they're going to be in. They want to treat you kind of like they did old ad networks, which is fine, I'll give you inventory, just don't tell them that you have this good inventory. If you remember the early days of ad networks. Whereas

• 10:46Speaker 01

  if you let the publisher sell it, they're perfectly happy to tell the advertiser, yeah, if you spend enough, you not only get a 30 second spot, but we'll actually put you in the show. I think that is the model I see in the US. That's the earlier model. And then the network model that took off internationally will start to enter the US.

• 11:03Auren Hoffman?

  These other markets are in the US is the best advertiser for these virtual product placement, CPG? Is that like clearly like the number one ICP?

• 11:14Speaker 01

  That's what we've seen in international markets. The CPG was the biggest, but at Rembrandt, we actually had luck because remember we started with creators. And so in cases like that, we were able to put billboard like stuff. And in that world, it doesn't matter. You could be a finance company. Actually, we saw movies being advertised inside this kind of stuff. Obviously, we saw CPG first.

• 11:38Auren Hoffman?

  Products like the new Pixel phone or something or whatever, or?

• 11:42Speaker 01

  Yeah.

• 11:43Auren Hoffman?

  So really, it could be anything that you would see on TV, you would see potentially here.

• 11:48Speaker 01

  Correct. If you're looking at 3D objects, it's by far CPG and auto. But once you start breaking out into, you know, using the billboard format, or sometimes indoors, we'd put a TV screen and we just put something on the TV screen. So those formats allow pretty much anything.

• 12:07Auren Hoffman?

  World of DAS isn't just a podcast. World of DAS membership is a private invite-only community for founders and executives of important data and AI businesses. It's a place for off-the-record, high-trust conversations focused on the biggest questions data leaders are facing. The community is about peer learning, curated intros, and access to private events. Everyone in the group is building or leading something meaningful in data. If that sounds like you, apply to join at worldofdas.com. That's world of D-A-A-S dot com.

• 12:48Auren Hoffman?

  One of the things that's interesting, and you and I have known each other for a while, you've kind of started some of these companies by using M&A as a lever. And I think that's super interesting. And that's maybe something like a first time founder can't do that well, but someone who's had a little bit more success like you could do. Can you walk us through that playbook?

• 13:07Speaker 01

  I think the first thing is that making acquisitions work is very hard. That's the first thing I would say. And one of the things that I learned at Oracle and at Cisco is they were really good with M and A. But I also saw mistakes in M and A. And the first number one cardinal sin of M and A is to walk in like a conquering hero, which is if you acquired a company, they do something you don't do, obviously, or else you would acquire them. And the temptation to scramble the eggs and immediately integrate them into the way you do work is exactly how you kill value. Of course, you need to find some integrated approach, but being able to understand what they do that's different, that you can preserve in a way that preserves their incentives to crush it is really the magic you have to get right.

• 13:51Speaker 01

  And that is like 90% of my focus isn't accelerating, making them look like us. It's protecting how they look different than us.

• 13:59Auren Hoffman?

  And I mentioned like Oracle is just going to have also a different way than, you know, if you're a smaller company, you're acquiring something. I mean, Oracle has a whole playbook and you can't argue with success. Like it's worked at Oracle. I mean, maybe it could work better, but.

• 14:12Speaker 01

  Yes and no. I'll tell you what was interesting about Oracle and Cisco. I had both of these experiences. So at Oracle, they took BlueKai, which was a DMP and a data exchange, and they split it in two. The DMP component was integrated into the Oracle Marketing Cloud with completely breaking it up into like seven pieces. And that piece of our business was growing over 100%. And after the acquisition, it went down way down. The other piece of our business, they didn't know how to integrate it. So they basically said, look, Omar, we don't want to mess this up. We're going to let you report into Thomas Kurian's org and he's going to let you acquire and we're going to let you run kind of independently. That one accelerated its growth. So I credit Oracle in allowing us to do that.

• 14:53Speaker 01

  And that's kind of what taught me this idea of, hey, we had a social experiment right in front of us. One piece got integrated too fast and the other one was more thoughtful. The thoughtful approach won by far.

• 15:05Auren Hoffman?

  The other thing is like interesting is these B to B brands like B to C are really important. Like the brand itself, the name is super important. When Procter and Gamble acquires a B to C company, they don't change the name to Productive PNG Soap. And so changing the name from BlueKai to Oracle Marketing Cloud is just like even 10 years later, people still call it BlueKai. Why change the name? Like, why do you have to do that? It's like immediately take out like 25% of the value like that.

• 15:36Speaker 01

  Totally. And similar thing happened at Cisco. We acquired a public company called IMI and they were essentially like the Twilio of Europe, but more enterprise and kind of integrated. And one of the things that we recognized is that Cisco people sell so real enterprise kind of deals. And IMI's approach was a usage-based model. And so if we had put it into the same salesperson's toolbox, they'd be like, I could sell a $7 million deal? Or you want me to sell this usage thing that I can't predict how much? So we kept it in some senses separate for a while.

• 16:11Auren Hoffman?

  Oh, so different sales teams. You didn't stuff the channel?

• 16:14Speaker 01

  We didn't stuff the channel right away. Eventually we figured out how to integrate it.

• 16:17Auren Hoffman?

  Yeah, or how to like incent people or whatever.

• 16:20Speaker 01

  Correct. But we took our time and, you know, may rest in peace, the CEO Jay Patel, who was acquired CEO, stayed with Cisco until he just recently passed away. Wonderful human being. And that group continued to grow and continued to be important to the acquisition thesis that we had. So that's kind of what I learned. And obviously in a startup, we can do these things more easily, which is if you're going to acquire something, figure out how to actually grow it and not suffocate it.

• 16:46Auren Hoffman?

  It is hard because sometimes you're acquiring like a real business with like a real P and L. And sometimes you're acquiring like an important feature with a great team. So there's different points we might want to integrate them in a different way.

• 16:59Speaker 01

  Yeah, so it's not an usual playbook to go out and do lots of acquisitions. At BlueKai, we just did one. At Voicia, we did one. Here at Rembrandt, we've also really just done one. It looks like we did two, but one was all cash and it was really just buying deals. We used our own technology and even the people, there were only two people. So that one, there was no equity involved.

• 17:22Auren Hoffman?

  Some of these deals, especially when you're a startup, they're like relatively small. Like, how do you even like find these companies to do it? And how does one like negotiate with them?

• 17:31Speaker 01

  Look, Spaceback, which we recently did, was a bigger deal to us. It was a profitable company with 3,000 advertisers. And the reason we found them is we had just launched a product in May with the Trade Desk. Actually, you'd relate to this. We sat down with the Trade Desk and said, look, decades ago, Jeff and I had sat down and said, how do we make sure that you can buy data as part of the bid for programmatic? And that was like the essence of our integration, Trade Desk, in the very beginning of the Trade Desk. And so we did the same exact thing here with Rembrandt and the Trade Desk, where we said, I want to be able to allow an advertiser to buy AI-generated creative without having to pay separate, where you literally, it's just part of your bid.

• 18:16Speaker 01

  And so they opened up a program, we got integrated in it, we launched. I think in June, they announced it. And the day of that announcement, they also announced another company in the same program called Spaceback. And we looked at their product and we're like, oh my God, they're doing the same thing with us, except without AI. It wasn't exactly the same. We would alter a video with AI. They would take a social post and augment it, add a QR code, change its format, animate the social component so it could run on CTV. But similar enough concept, and they were already profitable with 3,000 advertisers. So I'm like, oh my God, how much could we accelerate that business with AI? And we met the team and the team is like amazing in the sense that-

• 18:56Auren Hoffman?

  So they weren't running a process, you like created your own?

• 19:00Speaker 01

  Correct.

• 19:00Auren Hoffman?

  So you just like reached out to them? And because people aren't always receptive to that, like, how does one reach out to someone in the right way? Like, hey, guys, like-

• 19:09Speaker 01

  We started out with a partnership discussion and it just went deeper. And I think it was about common vision. So that's kind of how that played out.

• 19:17Auren Hoffman?

  You've run multiple companies. How does your like mature self run companies differently from your younger self?

• 19:24Speaker 01

  I'm gonna actually answer the kind of opposite flip side of that question. Cause there's a bunch of stuff that's obvious that you'll get right and better with. It gets easier to raise capital, it gets easier to hire people, it gets easier to understand you have to focus. And let go of you can't chase too many things, all those stuff gets easier. But I'll tell you some mistakes that don't get easier, that maybe even get worse. When you're a founder, the first one that gets worse is you forget this idea of a circle of competence. Which is that you really need to like. Be really, really good at something that helps you make decisions and understand things that other people don't understand. And that's your strength, but once you've succeeded a few times. It's very easy for you to overestimate your circle of competence. I'll give you an example of it.

• 20:09Speaker 01

  This one charity approached me and wanted me on the board. And I'm like, I really don't understand anything about your category. But they're like, oh, but you've been successful and this, that and that. And it's like, that's the common problem is that founders have been successful. People just assume that that success carries over to all these other areas.

• 20:30Auren Hoffman?

  Or other people might not question you. Like when you're the CEO, first time CEO, they're going to question you a little bit more than if you're the third time CEO.

• 20:37Speaker 01

  Yes, so that is like the temptation of arrogance. That's like the number one problem I would avoid. Another one that I get is like not losing your edge. I can't remember which of the Rockies movies it was, is where he was like. Training in really fancy places and then he got his butt kicked. And then his coach said, we need to go back to the basics. And he took him outdoors and had to chase chicken to remember how to get speed.

• 21:01Auren Hoffman?

  Oh yeah, I think that was Rocky IV.

• 21:03Speaker 01

  Yeah, and so that's kind of the thing I really like to remember is like, don't lose your edge. Just because it's easy to get capital, doesn't mean you should lose that frugality and have small teams. I'll give you an example, Dave Weiner, I've worked with him in all the last companies that have run. He's our chief product officer and our COO. He himself now still codes inside of Repli. And the reason is, is as a product officer, he wants to be able to prototype within the day. Something he wants and then hand it over to the engineering team and basically do lots of demos to say, let's get feedback before we scale the hell out of this feature.

• 21:43Speaker 01

  So getting people, no matter how successful, back into this gritty mode where they're closer to the customer, closer to the data, and not building teams that get in between them, teams and processes that scale things too fast before validation. It's like keeping that startup mentality is really crucial.

• 22:02Auren Hoffman?

  I recently had a chance to tour the Dallas Cowboys training facility and it is the nicest, most beautiful training facility you've ever been to. There's like, when you go to the auditorium, like the seats are like in Italian leather and they're gorgeous and they have like chefs everywhere and like the equipment is beautiful and polished and everything about it just looks like super nice. And so I was asking like the tour guide, I was like, how long have they had it, et cetera? And they were like, really explaining like, this is the nicest facility in the NFL. And I said, what is the worst facility, practice facility of all the teams in the NFL without a beach? He goes, Kansas City Chiefs is the worst one.

• 22:46Auren Hoffman?

  Okay, I think we have like a correlation, like they, right when this facility came online is when the Cowboys just started sucking. It's like, there's a point where it's just too nice. Like you need a little grit in your life.

• 22:59Speaker 01

  That's exactly it. You've nailed the core of the point. There's one more that's like so unexpected for me, which is false signal on product market fit. If you're a successful CEO, you've done this multiple times, you're going to find it easy to sell.

• 23:14Auren Hoffman?

  The first sales, you're going to make these sales so much easier, right? Do personal relationships.

• 23:19Speaker 01

  Yep, and renewal. And so you may be sitting on situations of renewals and think that you've got product market fit. And look, it can't be that you're selling something completely off. You probably are partially right. But you have to think differently as a second or third time to say, no, I really need to question, yeah, maybe it's good enough when I'm in the room, but does it scale when other people selling it? Have I really nailed the feature? That was unexpected to me. It was later in the process that I started to see that, no, we've got to fine tune this product. That's like when we started to get the insights that 3D objects alone in the room don't allow you to deliver a message. A billboard that stays on screen longer and

• 24:03Speaker 01

  takes over more of the screen gives you a new news message that fits the job to be done for the marketer. That was not obvious to me because the 3D objects were sexier and getting renewals. And so, things like that, where you literally, your presence and your previous success obfuscates things, and you need to be aware of that.

• 24:24Auren Hoffman?

  Interesting. Okay, so these are new mistakes that you make as a successful. What are some of the continued mistakes, mistakes you made before that you're like, I can't believe I'm making this mistake again. Like, what the heck? I've made this mistake now like five times in a row. Like, why have I not learned from this mistake?

• 24:42Speaker 01

  There are more simple things, like the way I run good team meetings or stuff like that got honed in over so many years by watching so many great other execs and learning from them, that sometimes later, I don't stick to the discipline of doing that. And it reminds me of, oh, I did this for a reason. Or planning process with a quarter, where I had a really good planning process and then, you know, I got too busy and I didn't do it as well as before. Those remind me to say, no, no, no, I can change it, or I can go back to it, but don't be less focused on the process of getting it right.

• 25:17Auren Hoffman?

  I have personally found that I make the same hiring mistakes over and over again. Like, I made this mistake 20 years ago, and then 15 years ago, and then 10 years. It's like, I can't believe I'm making this mistake again and again. Like, why have I not learned? It's almost like insanity. Like, you just be like, I can't believe I'm doing this. I don't know if you have something similar. You're a better CEO than me, so hopefully I don't have this.

• 25:38Speaker 01

  No, no, I think the hilarious thing about hiring, the one thing I learned about hiring is I have to surround myself by much more skeptical people than me, because I have this really bad attribute as a hirer, which is I want to like you. This is my default. Somebody comes in the room and I'm looking for reasons to like them. And so what happens with that mode is if my gut isn't a yes, it's a very clear no. But more importantly is I have people surrounding me in the interview process who provide a lot more skepticism. I hate to admit that itself. It makes me look like really silly when I admit that, but it's the truth. It's just my nature.

• 26:11Auren Hoffman?

  Okay, now you are known for running great team meetings, running effective board meetings. Like I remember 15 years ago, you turned me on to the Patrick Lencioni books. What are your advice to other people who want to be better at that?

• 26:26Speaker 01

  Whenever I forget this is when I suffer. I think there's three parts to a really good team meeting. And again, I think I learned this from watching. You probably remember this. At some point I did this tour interviewing and you were one of the interviews. Talking about the skills of really good CEOs and really good execs. And I take little bits and pieces from folks. So I essentially try to run a team meeting with three parts. The first part, and I tried to do this with a board meeting too, is what I call the Navy SEAL test, which is there has to be a very, very strong thread of accountability, which is I told you these things that I was gonna do. Here I am, and I don't get to change the ruler. I said these four things is what I need to hit. Here I am against those four things.

• 27:12Speaker 01

  So that's the Navy SEAL part, and it allows me to run a meeting where essentially I call it like the equal opportunity asshole. Like early in my career, it wasn't as easy for me to get into people's kitchen. And so I learned that by having a set of goals that I could push everyone on, I never felt bad. It wasn't personal. How are you doing on X? So that's the first part of the meeting is everybody's holding each other accountable previously set set of goals. And that way you know if you're making progress, you know what barriers you need to remove, you need to help people with. Great. That's part one. But if you only do that and you only have like metrics driven conversations, people are gonna tune out, except when they're talking about their metric. So you need the second part of the meeting, which is I call the suspense thriller. And that's where in advance

• 27:55Speaker 01

  you tell people a strategic topic that you want to get into. And it's called the thriller because nobody in the room knows how it's gonna end. And they want to participate to move the needle on it. And recently where I improved that part in dealing with a lot of Amazonians is it's great if somebody writes in advance.

• 28:15Speaker 01

  The depth on that topic you're going to discuss.

• 28:18Auren Hoffman?

  So you do the memo and then either people read it at the start of the meeting or they read it before the meeting or something or whatever?

• 28:24Speaker 01

  That's exactly right. For that portion of the meeting, which is the, hey, there's a strategic topic. We want to decide, are we doing this partnership or not? Are we changing our bonus structure?

• 28:33Auren Hoffman?

  Here's three options. Here's some pros and cons. Here's, you know, here's some questions you need to think about. Come prepared with this, like.

• 28:41Speaker 01

  Exactly. Again, I didn't, at the beginning, I had this right on the suspense thriller, but the addition of having someone do written good thinking in advance that everybody read, changed the quality of it.

• 28:51Auren Hoffman?

  Right. Cause sometimes you have the suspense thriller and then it's like, not everyone's up to speed on it. They can't even make the decision. They don't even have all the information. And then you spend the whole rest of the like, meeting, getting people up to speed on things.

• 29:03Speaker 01

  That's right. I had one CEO give me this critique early on when I didn't use this written format. He goes, you're okay with this very verbal culture. And I thought about what he said. And at first I felt a little bit, that doesn't sound very good. But then instead of being defensive, I'm like, I got curious. I'm like, okay, how do people do this right? And then we started adopting this mode. But if you just do a meeting with just this suspense thriller, even when done right, you lose the accountability because every week you could talk about something new and exciting, but then something may be broken in the business, you don't see it. So you need both. And then I got all those. And then the one thing that I was missing, which is you get into the metrics and you find what's going on, you have good strategic discussions, and a lot of these discussions

• 29:46Speaker 01

  are hard. You will need to address people's mental state. And Eric Rosa is the one that clued me into this because I watched how he did it. And he was so good at what I call the pep rally part. And a pep rally can't be fake. It has to be you identifying something that's going good and highlighting it and celebrating it. So people see a real example and feel good about it. And the engineer in me doesn't tend to do this. It's like, hey, let's just stick to the facts and move on. But as I observed the feeling of being in something like this with Eric, I'm like, that's a good skill. Let's make sure somebody on the team can do that.

• 30:27Auren Hoffman?

  It's actually really hard to do. I agree. Eric is like, he's a master at it. But most founders aren't that good at celebrating the wins.

• 30:36Speaker 01

  I found since I'm not that great at it, I need someone in my team who I know is naturally good. To lead that portion. And here's the thing is what I've found is I've honed this in after watching so many people, but for so many periods of time, you get too busy and you lose some parts of this meeting and you remember, oh yeah, let me stick to this. This is a good

• 30:57Auren Hoffman?

  These are the three parts. I love it. That's awesome. Not only do you run board meetings and stuff, but you also are a member of boards. You are a board member. You've been a member of a lot of great companies. How do you be a fact of board member?

• 31:13Speaker 01

  I think it's to really understand your contribution area and lead into it. For example, I was at LiveRamp for a while and I was very interested in how to productize as much as possible pieces of an org that had lots of service components to it. So that was like my swim lane, which is to lean into, okay, how do we productize that? And it was great because then, you know, your CEO knows what to come to you for and what to ask you to do. And you're not shy. You kind of lean in to that area.

• 31:51Auren Hoffman?

  And the other areas, you're more in like listen mode and stuff?

• 31:54Speaker 01

  If you looked at like the board at a LiveRamp, it had a bunch of folks that were super good at like finance. It had the whole range. Recently had Brian O'Kelly, who would dive into the engineering part. So you know your strength, you know their strength and it's important to do that. Otherwise, you either don't jump in enough because you don't feel like you know your swim lane, or you jump in too much, which is also

• 32:18Auren Hoffman?

  Yeah, that's not helpful.

• 32:19Speaker 01

  So it's finding where to go deep.

• 32:22Auren Hoffman?

  And speaking of LiveRamp, I am still a LiveRamp shareholder, even though basically the stock's been flat for 11 years. What hope can you give me over, you know, the next five years? I shouldn't just like cut and run. Again, it's like, at this point, it's totally personal. It has nothing. I just like love the company. So I don't want to sell, even though I probably should have sold a long time ago.

• 32:44Speaker 01

  The thing about them, let's first talk about what still remains to be kind of the core interesting thing about them. It's they have a dense network that includes like 225 third-party data providers, 500 media providers and CTV platforms and publishers.

• 33:03Speaker 01

  And what that means is that they've got audience access and identity access to 92% of the ecosystem. It means your match rates are going to be high. It means your measurements is going to be cross publisher. I think their value is in their network. And so how to make sure that turns into all the financial success you want is really the thing that's important. I think the thing that's been interesting to me lately, I mean, I left the board just two months ago, is they recognize this new wave that's going to come on the into advertising around Agentic. And they asked themselves, how can we play a fundamentally interesting role in that? Even if it doesn't provide huge revenue this year, but it grows.

• 33:48Speaker 01

  And they created this thing called the User Context Protocol, which is a piece of what you need in an Agentic world. And they actually donated it to the IB Tech Lab, because they really figured out that if this wave comes, they want to be the signal and the identity that is the underpinning of this Agentic advertising world. So I wouldn't bet on this being a big driver of revenue next quarter, but it's good to see them thinking through a wave that they can be fundamentally weaved into.

• 34:19Auren Hoffman?

  If you think of like the two ad tech companies that have been most successful, one of which you're on the board of, you've got the Trade Desk and you've got Apple Lovin'. Why did they do so well? Why did those two companies, they did an order of magnitude better, sometimes maybe even two orders of magnitude better than everybody else? Like, what did they do so well?

• 34:39Speaker 01

  Yeah, and they both do not own their own Facebook or Google or Walgreens.

• 34:44Auren Hoffman?

  That's right, they're not O and O. Apple has a little bit, but not to the same extent.

• 34:49Speaker 01

  And so that makes it even harder. I think two things that stand out for both of them for me. Number one is laser focus on what they do from the very, very beginning. In the case of Jeff, you listen to him now and you listen 10 years ago. He's always been about creating a

• 35:06Auren Hoffman?

  For our audience, it's Jeff Green, the CEO of the Trade Desk.

• 35:09Speaker 01

  Yes, that's right. Always been about creating a fair, transparent auction for the open internet. It hasn't changed. And when you sit down and talk to him and go five levels deep. It's not like fluff. He really means it and he will go really deep into aspects of the auction that will make it either more fair or more open. And he's even had conversations where, you know, before I was on the board, he'd say, look, there are areas where we pushed so hard for transparency, maybe we made it too complex for some people. So recently we had to do these other things to say, well, people also wants it simple. It's just that depth of awareness. I rarely see that in CEOs of ad tech where they've stuck to the same thing and just kept going deeper. But the thing has to be fundamental. You can't stick to the same thing for 10 years

• 35:54Speaker 01

  if it's not fundamental enough. So that's one attribute. The other attribute of both of the companies that's interesting is they're both buy side. They're not sell side and the dollar in advertising flows from the advertiser. You know, you see a lot of companies that did variations of sell side things that just didn't receive the scale because everybody comes in and says, oh, your middleware and, and they get squeezed. So

• 36:17Auren Hoffman?

  Getting as close to the money is where the money's coming in, not where the money's going out is better.

• 36:22Speaker 01

  Exactly. That's what I see from them. Also, if you look at the trade desk that I found interesting is day one, everybody else in ad tech really early on were like, oh, I don't like agencies. They keep churning and they're hard to deal with. And the trade desk embraced the agencies.

• 36:37Auren Hoffman?

  Yeah, they made a bet that they were going to be the agency's friends.

• 36:40Speaker 01

  Correct.

• 36:41Auren Hoffman?

  It's interesting because both Jeff and Adam, the CEOs of the trade desk and Apple and respectful respectfully, like they are both really deep product people, like they're really good product people. I mean, they're obviously good at sales and you know, these other things that are important, but I'd say product is where they shine. And historically, like the tech companies run by really good product people are the ones that make the most money.

• 37:08Speaker 01

  I'm totally with you. I tend to have that bias of I prefer, that's what I grew up in is the product side. And I'll tell you, like, the one company that has always surprised me is that how Apple, who essentially run by an ex CFO, has done so well over so long. That's a counter example. But if you look at like Microsoft, when Satya came back, he was a product CEO and that really, really served Microsoft well. So that's my bias.

• 37:34Auren Hoffman?

  Yeah, and it's interesting, like, the ad tech category, I remember like when you and I were both younger in the category, I used to bifurcate the companies. There's companies that were on the ad side and there's companies on the tech side. And the tech side just like was a better business long term than the ad side.

• 37:52Speaker 01

  In the ad side, people just keep coming back to scaling the relationship sales part of things.

• 38:00Auren Hoffman?

  Steak dinners and, you know, Yankee tickets and stuff.

• 38:03Speaker 01

  Which is, yes, you will get scaled that way, but it won't give you the same financial performance over the long run.

• 38:09Auren Hoffman?

  What are the big, massive opportunities in ad tech today that you think people are sleeping on or they're not yet taking advantage of?

• 38:18Speaker 01

  I'll answer the, what are the big ones? I think people are aware of these now. But the area, like, for example, that we're focused on, and even broader than what we're specifically doing, is the idea that the creative itself, rather than figuring out how to take a piece of creative that's done, and figuring out how to bid on it and where to place it, how to price it and what data you use targeted, that gives you like incremental efficiency. But changing the message itself and how it gets delivered and what it looks like in a way specific to this user, that's massive. And only AI has made this possible. And this is going to be an area that's going to continue to get investment.

• 38:59Speaker 01

  In the past, companies that went after things like DCO, it was very hard for them to have a very good outcome. But now, AI, the whole idea of opening up the creative itself is, I think, a very interesting area that's just going to continue to get a lot of attention. And I'm investing in some companies that use essentially AI to generate organic posts. And what you find is the part of the market that that's going to work in, and you also have ones that do in advertising, are going to be SMB. If you're a top 10 company, you're going to want everything to be so pixel perfect. You're going to constrain the hell out of this stuff too much and you can't learn, have data and volume on that kind of constraint.

• 39:42Speaker 01

  Whereas SMBs make it more easy and if you.

• 39:45Auren Hoffman?

  Yeah, the local coffee shop or whatever is perfect.

• 39:47Speaker 01

  Correct. And then it's going to get better and better and better. And then finally it'll break the top category. So I think, you know, getting into the creative part is an area that's interesting. And then the other area that everybody's jumping on now, which is kind of agentic. It's not area people are asleep. There's a lot of hype and the hype is ahead of the revenue. But nevertheless, it'll still be an interesting innovation that will disrupt the ecosystem. So I don't know how.

• 40:11Auren Hoffman?

  What do you mean by the agentic? What's the thing?

• 40:13Speaker 01

  All the different things around agentic media buying with, you know, ad CP, what LiveRamp did with the user context protocol. And just even rethinking all your interfaces to be agentic, being able to have any interface and have a conversation with user. And actually change the product that you show the user based on a deeper intelligence inside of the agent.

• 40:35Auren Hoffman?

  On the first thing you mentioned, one area that still puzzles me why it's so bad is just generally recommendation engines. So if you think of like an ad, ad is really just a recommendation of content. Today, even when I want to get good recommendations on Netflix, on Spotify, on Amazon. The recommendations that they give me internally are pretty bad, even they have a ton of data. I mean, just think of all the Spotify things they have, all the, it's just so bad. I'm still puzzled why nobody's been able to make great recommendations. So I don't know if you've thought it through.

• 41:13Speaker 01

  Well, I mean, first of all, I had a recommendation company at once back in 2002 and we actually ran Barnes and Noble recommendation engine and Nordstrom and Tower Records.

• 41:23Auren Hoffman?

  I don't know if your recommendations are any better than it was in 2002. The collaborative filtering thing, whatever you did back then may not be any different today.

• 41:31Speaker 01

  No, that's the point I was going to make, which is the vast majority of all recommendation engines. Still have that people who bought this bought that core. And the reason is, is the information density at the population level. Is so much better than the information density at the individual level. At least that's how it was back then. So people just for compute reasons just said, it works good enough, just do this and it gives a lot of sales lift. And I don't think most of the recommendations that people see are actually personal. They're more at the level of titles. I think Netflix might be an example, a counter example of that. And Amazon, they do get to the personal level in the recommendations.

• 42:11Auren Hoffman?

  Oh, really? You think I like, I still find them so like, if I say, show me some shows I want to watch, or show me some things I want to buy. It's pretty rare that I want any of those things. Like, I'm so surprised how bad it is. Like, it would be amazing. I would just like literally go buy stuff on Amazon if it just showed me some things I want to buy. But it's so terrible.

• 42:31Speaker 01

  I think the area where recommendations can change and should change is they should start using the wisdom available to L L Ms. For example, today, we're in November. There's a certain temperature in San Francisco. If you look at my past purchases, but also filter by the temperature in November, the fact that there's a bunch of trending fashions that are coming around us in the next couple of weeks. Those should go into it and L L Ms can take that into account. Whereas data engines that are looking for data density may be using data up to two years old. Because that's where the data density is.

• 43:09Speaker 01

  And so we need a different approach that uses less data and more contextual understanding of today.

• 43:18Auren Hoffman?

  Yeah, because you think of an ad like an ad is really, it's just a content recommendation. Ideally, it's showing you real content that you want to see. And some people call them ads, but some people like actually pay to get recommendations. Like you would pay a travel agent to show you great recommendations of places you should go to. You would pay an interior designer to show you great recommendations of things to buy. Like, that's what they do. They're showing you ads and you pay them a lot of money to go do it. That's what we want. I just want to see ads that are like relevant to me.

• 43:48Speaker 01

  Totally. And what's really good about someone, when you ask someone for a good recommendation of dinner, by the way, is they'll ask you a bunch of questions. They'll start by really understanding what you like. An approach that says, let me figure out how you're, instead of asking you a question, let me just assume you're similar to these million other people and get their answers. That's the problem.

• 44:08Auren Hoffman?

  You're also on the side now are part of Venture Studio at Superset. And if you think of VC firms, basically almost every VC firm is structured the exact same way. But pretty much every venture studio, if there's 100 venture studios, they're structured in 100 different ways. Why is that?

• 44:27Speaker 01

  I think that if you look at a venture studio like Superset is the first thing we say is we're not really VCs. We're not attempting to be VCs. Essentially what we are is we are builders. So you look at Tom and Vivek have built multiple companies and sold multiple companies. Same thing here, Jeremy, Peter, we're operators. And we want to scale our ability to start and build great companies. It's very different than VC, which is they're trying to scale their ability to invest in other people's companies. So then if you look at how would a group of people like Tom and Vivek and myself and Jeremy and Peter and others, how would we build companies together?

• 45:09Speaker 01

  It's based on kind of similarity in our cultural values. And so that's why it would look different than another venture studio that thinks in a fundamentally different way. Our approach is to say, when we invest, either create a new company and bring in a CEO or a CEO comes to us, we're on both sides of their cap table. We're both like a co-founder to them and also a source of money. That's what makes it a bit different.

• 45:33Auren Hoffman?

  What types of companies do you think are better suited to like a studio model versus more traditional model?

• 45:42Speaker 01

  In a traditional model, you want the money and you want to see them once a quarter. In this model, if you want that team to help you accelerate better decisions and operate better, you're almost open to almost the level of a co-founder, it's not quite that much, but.

• 45:58Auren Hoffman?

  I thought you generally are the co-founder. I mean, you're kind of like helping found the company, you're helping come up with the idea, right? Sometimes you are coming up with the idea and then you're convincing someone to be the founder.

• 46:08Speaker 01

  Exactly right. So you're gaining a co-founder and we have different styles, which is sometimes we completely start it and bring in. And other times it's someone who's kind of heading a direction and they bring us as a co-founder is the way of thinking. So that immediately filters out certain types of companies. Like if you're a repeat founder and know exactly what you want to do and you just want money, maybe not the right fit.

• 46:27Auren Hoffman?

  But is there a particular type of company you think is better for one or the other? Or do you think it like, if you're going after certain markets or you think it just kind of does just depend?

• 46:37Speaker 01

  We tend to favor companies that have a data and AI focus. That's something we tend to focus on. One of the companies that participating in, I won't say too much about them, other than saying, first time founders, really young, super hungry and deep in the area they're in, but enjoy having ad tech operators helping them figure out stuff. So that kind of match was like chocolate and peanut butter. Whereas, you know, like I said, I met a repeat founder and knew exactly what they wanted to do and just wanted money. I'm like, yeah, maybe that's a different path. Maybe you want to go classic VC.

• 47:13Auren Hoffman?

  I've helped companies where we're kind of founding it together and we realize actually all they want from us is money. You know, it's value added you can bring, but it's not as great if you can also bring some other strategic things to the table.

• 47:26Speaker 01

  Yeah, and you may even do some of those, but at least you do it knowing that that's what it is.

• 47:30Auren Hoffman?

  Now, Tom Chavis was one of the founders of Superset. He was your competitor. He was the CEO of Crux when you were at Blue Kai. I mean, literally you couldn't have been more competitive, these two companies. I assume that somehow brought you closer in some way, or maybe you hated each other before and now you love each other. Like, how did that relationship evolve?

• 47:50Speaker 01

  Look, we were very competitive. Some of our team members like were so competitive, they just like really disliked each other. Tom and I didn't have the dislike part. We just knew we were competitors. And then after it, when I was aboard LiveRamp, we took interest in Habu. And, you know, we ended up acquiring Habu.

• 48:06Auren Hoffman?

  And you buy, bought the company?

• 48:07Speaker 01

  Yeah, and after that, I spent time with Tom and I really got to like him and the feeling was mutual. And then, you know, because of that context, they invested in Rembrandt. And through that, they said, hey, can you also become a GP? But because we know you're running Rembrandt and you're busy, our time requirement's going to be, we're going to honor the fact that you're a CEO. So we don't expect you to be full time like the rest of us. So it kind of just worked out great. What's interesting about Tom, Vivek, Jeremy, these guys is they have the most important thing I care about with execs, which is the combination of competence and low ego. I don't work great with high competence, high ego. These guys are very thoughtful people.

• 48:49Auren Hoffman?

  What does it mean to be low ego? Like, what does that actually mean?

• 48:53Speaker 01

  Essentially, there are part of, I think, like, if I could just be honest, part of the Western culture on success and what you see in the movies is super high. I'm just fucking right and I'm gonna pound the table and everyone's gonna do what I want. That alpha male image, I think, is a destructive image because the two attributes are entirely unrelated. You can be very deep and confident in what to do without being arrogant and you could be very arrogant and dumb.

• 49:27Auren Hoffman?

  Okay, so ego is more arrogance in this case?

• 49:30Speaker 01

  Yes.

• 49:31Auren Hoffman?

  Not like thinking highly. I assume all these people think super highly of themselves, but it's the arrogance part that's the problem.

• 49:38Speaker 01

  High self-worth is a good thing. Arrogance mixed in with a high ego, which is like, for instance, if you challenge someone, the understanding that they're challenging the context, not the person. A high ego person doesn't accept that and they have to be right. And they need you to lose for them to win. And these guys are not like, we will debate issues and none of it is personal. And so it's very helpful to have that kind of environment.

• 50:05Auren Hoffman?

  All right, this has been amazing. Last question we ask all of our guests. What conventional wisdom or advice do you think is generally bad advice?

• 50:12Speaker 01

  I'll start with this one thing that has served me all of my life until recently. Is this idea that to succeed, just focus on what you can control and get better and better and better at it. And take no excuses. Any failure is your fault and you just need to try harder. And that has served me so well all of my life. Until at one point, I had one daughter who had a learning disability and I realized. That I had been driving her hard, not knowing that this is not a problem of she's not working hard enough. She literally has a serious learning disability that

• 50:53Speaker 01

  once we identified it, she was able to get the right accommodations for it. And then, her grades started getting better. She ended up going to Berkeley. Then she went to grad school, like, everything turned around. And I was like, oh, my God, my worldview was lacked such empathy. It always saw not succeeding as a sign that you're not trying hard enough.

• 51:14Auren Hoffman?

  And by the way, I'm glad you have a towel call at Berkeley because you've had too many Stanford people in your family. And it's a big problem. So I'm glad you've got some Berkeley people, too.

• 51:24Speaker 01

  But anyway, she taught me so much about a worldview that only focuses on succeeding and trying harder and having no empathy for people who have struggles.

• 51:34Auren Hoffman?

  That's a good one. You said you have another one as well?

• 51:36Speaker 01

  No, not that same thing, but that lack of empathy in my worldview is something that needed to be addressed.

• 51:43Auren Hoffman?

  This has been amazing. Thank you, Omar Tawakol, for joining us on World of DAS. I follow you on LinkedIn. We've been friends for a long time. Huge fan of yours. I definitely encourage our listeners to engage with you. This has been really interesting and a ton of fun.

• 51:56Speaker 01

  Thanks for having me, Aur. And I've been a huge fan and followed your success. And we love what you're doing at Flex Capital. So thanks for having me.

• 52:05Auren Hoffman?

  Thanks for listening. If you haven't already done so, please subscribe to World of DAS wherever you consume your podcasts, YouTube, Apple, Spotify, and more. And please help us get discovered by leaving a review and check out worldofdas.com. That's worldofdas, D-A-A-S, dot com. And of course, connect with me on Twitter, at Aur. That's A-U-R-E-N. Would absolutely love to hear from you.